WO2020132766A1

WO2020132766A1 - Method for dimensioning a wdm optical network with wavelength continuity constraint

Info

Publication number: WO2020132766A1
Application number: PCT/CL2019/050122
Authority: WO
Inventors: Reinaldo Antonio VALLEJOS CAMPOS; Nicolás Alonso JARA CARVALLO
Original assignee: Universidad Técnica Federico Santa María
Priority date: 2018-12-28
Filing date: 2019-11-26
Publication date: 2020-07-02
Also published as: US11368769B2; CL2018003860A1; US20220078533A1

Abstract

The invention relates to a new method for jointly defining a policy for assigning wavelengths to each network connection and for calculating the number of wavelengths in dynamic WDM optical networks without wavelength conversion. To solve this problem, the method comprises including in each network connection a fixed route for transmitting, which is defined before operating the network. This new approach has two main differences from previous strategies.

Description

METHOD FOR DIMENSIONING A WDM OPTICAL NETWORK WITH WAVE LENGTH CONTINUITY RESTRICTION

TECHNICAL FIELD OF THE INVENTION

This invention relates to the field of optical networks, in particular with WDM optical networks with wavelength continuity restriction and specifically provides a method for calculating the capacity of a WDM optical network with wavelength continuity restriction.

BACKGROUND OF THE INVENTION

A central problem in WDM optical networks is knowing how many wavelengths should be assigned to network links. Finding the number of wavelengths for the bonds is called the "Wavelength Sizing" (WD) problem. The number of wavelengths significantly affects the cost of the network, as it determines how many infrastructure resources are needed on the network to achieve the desired network performance. Therefore, efficient sizing of each link in an optical network is of utmost importance.

To achieve efficient wavelength sizing in dynamic networks, two contradictory objectives must be met. On the one hand, it is desirable to decrease the cost in infrastructure and, therefore, the cost of the network; on the other hand, it is necessary to guarantee a certain level of Quality of Service (QoS) to network users, measured by the probability of user blocking. In other words, the network is designed to offer users a very low probability of blocking (a value close to 0) while saving significant network resources. Solving this problem in WDM dynamic optical networks is especially difficult when the network does not have wavelength conversion capabilities. This lack of wavelength conversion means that when a connection wants to transmit, the same wavelength must be available on every link that belongs to the given connection path. (end to end). This is known as the wavelength continuity constraint.

To solve the wavelength dimensioning problem in dynamic networks without wavelength conversion, several studies have been proposed so far. For example, the book "Optical networks: a practical perspective", by Ramaswami et al., 3 edition. Morgan Kaufmann Publishers Inc., San Francisco, CA, USA, 3rd edition, 2009, presents a collection of such studies. Among them, homogeneous sizing is the most common method used to date. However, this procedure produces an oversize of the network and, consequently, a loss of efficiency.

To solve the wavelength dimensioning problem, you need to define what wavelength each connection (also called "user") will use each time they want to transmit, denoted as the "Wavelength Assignment" (WA) problem . It should be noted that the chosen wavelength changes over time, as each time the user requests resources to send data, the WA method searches for an available wavelength on the user's path links.

This problem has been extensively covered by previous work. Some of the most common heuristics that are proposed in the literature to solve the WA problem are: First-Fit (FF), Random Fit (RF), Most-Used (MU) and Less-Used-Use (LU), among others. . First Fit is the most common and fastest method used to date, as can be seen in the article "Routing and wavelength assignment analysis in large WDM networks", by RT Koganti and D. Sidhu, in Procedía Computer Science, vol . 34, pp. 71-78, 2014.

A long-standing problem in optical networks has been to provide a certain quality of communication to network connections, despite the existence of different classes of users. This multi-class traffic occurs due to the existence of different priority users (or quality of service requirements), or some mechanisms to intentionally eliminate non-compliant bursts to improve network performance. However, the usual wavelength allocation and dimensioning approach (First-Fit wavelength allocation with a homogeneous dimensioning procedure) does not take into account these problems. For example, the first tuning wavelength mapping scheme does not take user classes into account, as it simply searches sequentially for an available wavelength, without making any class distinctions. On the other hand, in homogeneous sizing, the number of wavelengths assigned to network links is usually defined by the user class with the strictest quality of service (QoS) requirement, which gives classes a priority lower, better quality of service than requested in your SLA.

US Patent 9,060,215 describes a genetic algorithm for solving routing and wavelength assignment in optical networks with dynamic wavelength routing. The method proposes an iterative method, based on genetic algorithms. The network operator can choose the fitness function to find a good solution based on any criteria defined by the network operator.

US Patent 8,693,871 describes a method of solving routing and wavelength assignment. This method seeks to minimize the number of wavelengths used, based on a solution of container packaging problem, seeking to solve the minimum number of containers, where the wavelengths are containers. This method assigns wavelengths with the new versions in the First-Fit, Best-Fit, First-Fit Decrementing and Best-Fit Decrementing techniques through the use of containers. This method does not take into account the quality of service offered to each user and uses a homogeneous sizing algorithm.

SUMMARY OF THE INVENTION

The present invention provides a method for implementing a WDM optical network with wavelength continuity restriction that is based on a definition of a policy to assign wavelengths to each network connection, and on a calculation of the number of wavelengths. wave in WDM dynamic optical networks with wavelength continuity restriction. The present invention provides a method for implementing a WDM optical network with wavelength continuity restriction characterized in that it comprises the steps of:

obtaining, in a processor, the topology of said optical WDM network, said topology including nodes and links in said optical network; obtaining, in said processor, a set of connections in said optical WDM network and a set of threshold values for each connection of said set of connections;

obtaining, by means of said processor, a set of routes that allow satisfying said set of connections in said topology of said optical WDM network;

determining, by means of said processor, a set of numbers of wavelengths for each link of said topology, wherein said number of wavelengths is determined in such a way that, for each connection of said set of connections, a probability value of blocking corresponding to said connection is less than the threshold value corresponding to said connection; and

implement said optical network according to said set of wavelength numbers.

In a preferred embodiment, the method is characterized in that said set of wavelength numbers is determined by the steps of:

start each link with a number of wavelengths equal to 1; determine, by means of said processor, the probability of blocking for each connection;

determining, by said processor, for each link, if the blocking probability corresponding to each of the connections using said link is less than the threshold value corresponding to said connection; increase by 1 the number of wavelengths corresponding to a link if there is a connection that uses that link where the blocking probability is greater than its corresponding threshold value; and iteratively repeat the previous three steps.

In a more preferred embodiment, the method is characterized in that it additionally comprises storing the number of wavelengths required for each connection of said connection set. In another more preferred embodiment, the method is characterized in that it additionally comprises storing the number of wavelengths required for each link of said optical network.

In a preferred embodiment, the method is characterized in that said set of routes is determined using the Dijkstra algorithm.

In another preferred embodiment, the method is characterized in that, to determine said blocking probability, said processor executes the steps of:

define a plurality of layers, each layer corresponding to the topology of said optical network, where the number of wavelengths available on each link is 0 or 1;

determining, for each link in said optical network, the connections using said link; define, for each connection using said link, a first numerical value, t _{0N c} , corresponding to the time in which said connection is active and a second numerical value, t _{0FF c} corresponding to the time in which said connection is inactive ; define a plurality of corresponding values Bu _{blocking probability to the connection on the link c in layer i wy obtain a value _C corresponding to BP blocking probability connection network c.

In a more preferred embodiment, the method is characterized in that for said steps of defining a BL _{ value and obtaining a BP _C value they are performed iteratively. BRIEF DESCRIPTION OF THE FIGURES

Figure 1 shows an input and output diagram of a first embodiment of the method that is the object of the present invention.

Figure 2 shows a flow diagram of a first embodiment of the method that is the object of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

Hereinafter, the invention will be described in detail, with reference to the figures that accompany the present application.

The present invention provides a method for implementing a WDM optical network with wavelength continuity restriction, essentially comprising the steps of: obtaining, in a processor, the topology of said WDM optical network, said topology including nodes and links in said optical network; obtaining, in said processor, a set of connections in said optical WDM network and a set of threshold values for each connection of said set of connections;

implement said optical network according to said set of wavelength numbers. As illustrated schematically in Figure 1, the method that is the object of the present invention, in a preferred embodiment and without this limiting the scope of the present invention, can be understood, in terms of block diagrams, as a block whose inputs are the topology of the network, represented as a pleasant Q = (, L) where is the set of nodes and L is the set of links in the network; the traffic load, Q _c , for each connection c defined in said optical network, and a set of pre-calculated routes 31, which allow to implement the connections, c. Additionally, are the values considered as input to the method? _c , where? _c is the maximum blocking probability for connection c, and corresponds, interchangeably, with the threshold value corresponding to connection c. As block outputs, in this preferred embodiment and without this limiting the scope of the present invention, the number of wavelengths per link is obtained,

and the maximum number of wavelengths for each connection, u _c . The method is schematically and without limiting the scope of the present invention, as a "Fair Assignment Heterogeneous Wavelength Sizing" (FairHED) method.

Said block can be implemented, for example and without this limiting the scope of the present invention, by means of a processor configured to carry out the method that is the object of the present invention. The way in which said processor obtains and executes the instructions that allow it to execute the method that is the object of the present invention does not limit the scope of the present invention. For example, and without this limiting the scope of the present invention, said instructions may be stored in a storage device operatively coupled to said processor. Said storage device may be internal or external to said processor without this limiting the scope of the present invention. However, in other exemplary embodiments, said processor can obtain said instructions remotely, for example and without this limiting the scope of the present invention, by means of a local area data network, or a wide area data network. , just like the internet. In an exemplary embodiment, without limiting the scope of the present invention, said instructions may be encoded in a ad-hoc programming language, and said processor may be configured to compile and execute said instructions. However, in other exemplary embodiments, said instructions may be in a format executable by said processor, and said processor may be configured to execute said instructions.

Hereinafter, without limiting the scope of the present invention, it will be understood that the steps of the method that is the object of the present invention are implemented by said processor, unless it is explicitly indicated, or it is understood from the context, that they are implemented by another device.

The method that is the object of the present invention comprises the step of obtaining, in said processor, the topology of said WDM optical network, said topology that includes nodes and links in said optical network and that, as mentioned above, can be represented, without that this limits the scope of the present invention, by means of a graph Q = (J \ f, X) where J \ f is the set of nodes and X is the set of links in the network; and the step of obtaining, in said processor, a set of connections in said optical WDM network and a set of threshold values for each connection of said set of connections. Illustratively, without limiting the scope of the present invention, said connection set will be denoted as set X and each connection in said connection set as connection c. On the other hand, said set of threshold values, as previously mentioned and without this limiting the scope of the present invention, correspond interchangeably with a maximum probability of blocking the connection c and are denoted as? _c. The way in which said processor obtains said topology, said set of connections or said set of threshold values does not limit the scope of the present invention and will depend on the specific way in which the method that is the object of the present invention is implemented. For example, and without this limiting the scope of the present invention, said processor may be operatively coupled to a graphical interface that allows defining said topology of said optical network, as well as said connections c and said threshold values b ₀ . However, in other preferred embodiments, said processor may obtain said objects by means of a set of data representing said topology of said network, said set of connections or said threshold values, b ₀ , without this limiting the scope of the present invention. On the other hand, as previously indicated, in a preferred embodiment, the set of routes R is an input of the method that is the object of the present invention. However, in other preferred embodiments, without this limiting the scope of the present invention, said processor may be configured to obtain said set of R routes from the topology of the optical network to be implemented. To obtain said set R, said processor can make use of any method known in the state of the art, such as, for example, and without limiting the scope of the present invention, by means of the Dijkstra algorithm.

The method that is the object of the present invention further comprises the step of determining, by means of said processor, a set of numbers of wavelengths for each link of said topology, wherein said number of wavelengths is determined in such a way that, for each connection of said connection set, a blocking probability value corresponding to said connection is less than the threshold value corresponding to said connection. The manner in which said processor determines said number of wavelengths corresponding to each link does not limit the scope of the present invention. In a preferred embodiment, without limiting the scope of the present invention, said set of wavelength numbers is determined by an algorithm comprising the steps of: - starting each link with a number of wavelengths equal to 1; determine, by means of said processor, the probability of blocking for each connection;

determining, by said processor, for each link, if the blocking probability corresponding to each of the connections using said link is less than the threshold value corresponding to said connection; increase by 1 the number of wavelengths corresponding to a link if there is a connection using that link where the blocking probability is greater than its corresponding threshold value; and

iteratively repeat the previous three steps.

A preferred embodiment of said algorithm, without this limiting the scope of the present invention, is illustrated schematically in Figure 2. First, the processor initializes a set Q (10) that stores all those connections for which it is You have satisfied the condition that your blocking probability is less than your threshold value. As is evident for a person with average knowledge in the technical field, said set Q is initialized empty, since no connection has satisfied this condition.

After initializing said set Q, and because each link i has a wavelength value equal to 1, the probability of blocking BP _C for each connection, c (20), is calculated. The way in which said blocking probability is calculated does not limit the scope of the present invention and will depend on the specific application of the method that is the object of the present invention. However, by way of illustration and without limiting the scope of the present invention, it is understood that the exact calculation of said blocking probability involves a high computational cost, therefore any way of estimating said blocking probability can be implemented without this limits the scope of the present invention and the terms estimate said blocking probability and calculate said blocking probability will be used interchangeably.

In a preferred embodiment, said blocking probability is calculated using the algorithm published by Jara, et al in the article “Blocking Evaluation and Wavelength Dimensioning of Dynamic WDM Networks without Wavelength Conversion”, published in Journal of optical Communications and networking, Piscataway, NJ ; Washington, DC: IEEE: Optical Society of America, 2017, 9 (8), pp. 625-634. Generally, without limiting the scope of the present invention, said method for calculating said blocking probability comprises the steps of:

define a plurality of layers, w, each layer that corresponds to the topology of said optical network, where the number of wavelengths available on each link is 0 or 1;

determining, for each link in said optical network, the connections using said link;

define, for each connection using said link, a first numerical value, t _{0N c} , corresponding to the time in which said connection is active and a second numerical value, t _{0FF c} , corresponding to the time in which said connection is inactive; define a plurality of BL _{ values corresponding to the probability of blocking connection c on link í, in the wy layer - obtain a value BP _C corresponding to the probability of blocking connection c in the network from said values BL.

In this method, without limiting the scope of the present invention, the traffic load corresponding to link c, Q _c , is calculated using the formula: t () N, c

Qc = - toN, c + t () FF, c In a more preferred embodiment, without this limiting the scope of the present invention, said values t _{0N c} and t _{0FF c} correspond, for different connections c, to identically independent random variables distributed (iid).

Moreover, in another preferred embodiment, the value BL ^ _{is calculated by the formula:

where the values

and f ™ are defined as follows:

with to _FF .c ^the time in which connection c remains inactive in layer w; and

with Tf the set of connections c using link l, provided that the number of wavelengths available on link i is equal to 1 in layer w.

Then, from these BL _{ values, the blocking probability of connection c in layer w can be obtained, using the formula:

with r _c being the route that allows connection c, which allows obtaining the probability of blocking connection c in the network using the formula:

In a more preferred embodiment, the calculation of the blocking probability is performed iteratively. More particularly, without this limiting the scope of the present invention, said iterative calculation is performed by iteratively repeating the steps of: defining a plurality of BL values corresponding to the probability of blocking of connection c on link i, in the layer w and obtain a BP value _C corresponding to the probability of blocking connection c in the network from said BL values. To obtain a convergent iteration, once the algorithm has been implemented for the first time, the t _{gFF c} values are updated using the following formulas:

After said blocking probability, BP _C , has been found for each connection c, said BP _C values are compared with the corresponding threshold values of connection c (30), provided that said connection c is not in the set Q (or, equivalently, that said connection c is in the set X \ Q). If, for a specific connection c, is it true that BP _C is less than the threshold value? _c , said connection c is stored in the set Q and the number of wavelengths available in the links that are part of the route r _c is stored as the value u _c .

Then, for each link, í, it is evaluated whether the connections using that link have satisfied the condition BP _C <? _c (40). If there is any connection that has not satisfied the previous condition, the number of wavelengths available on said link is increased by 1. Subsequently, it is verified if the set Q is equal to the set X (50), that is, it is verified whether is that all connections c have satisfied the condition BP _C <? _c . If there is any difference between said set Q and set X, the steps of calculating the probability of blocking BP _C are repeated for each connection, c (20); comparing said BP values _C with the corresponding threshold values of connection c (30) and storing in the set Q those connections that have satisfied said condition; evaluate whether the connections using each link, i, have satisfied the condition BP _C <? _c (40) and increase the number of wavelengths available on a link if there is any c connection that uses that link and has not satisfied its condition; and verify if the set Q is equal to the set X (50). On the other hand, if it is verified that the set Q is equal to the set X, that is, that all the connections have satisfied their operating condition, the number of wavelengths available for each link, W _f , and the number are stored of wavelengths required for each connection, u _c (60).

In this way, the method that is the object of the present invention presents a series of advantages with respect to other methods of implementing optical networks known in the state of the art, without this limiting the scope of the present invention.

First, the method naturally assigns a different number of wavelengths to each link in the network (which is commonly referred to as heterogeneous sizing). The criterion for assigning the number of wavelengths on each link is that the probability of blocking of each connection does not exceed a threshold value which can be, for example and without limiting the scope of the present invention, defined in an Agreement Level of Service (or SLA).

Secondly, the method, naturally, assigns to each connection a number of wavelengths that allows the blocking probability of said connection to be as close as possible to the corresponding threshold value, thus avoiding network oversizing.

Thirdly, as a consequence of the above, the method naturally allows the network to be dimensioned to serve connections with different threshold values, that is, the method allows solving the network dimensioning problem when there are classes of users with different level agreements of service.

Claims

1. A method to implement a WDM optical network with wavelength continuity restriction, CHARACTERIZED because it includes the steps of:

obtaining, in a processor, the topology of said optical WDM network, said topology including nodes and links in said optical network;

obtaining, in said processor, a set of connections in said optical WDM network and a set of threshold values for each connection of said set of connections;

determining, by means of said processor, a set of numbers of wavelengths for each link of said topology, wherein said number of wavelengths is determined in such a way that, for each connection of said set of connections, a probability value of blocking corresponding to said connection is less than the threshold value corresponding to said connection; and implement said optical network according to said set of wavelength numbers.

2. The method of claim 1, CHARACTERIZED in that said set of wavelength numbers is determined by the steps of: starting each link with a number of wavelengths equal to 1; determine, by means of said processor, the probability of blocking for each connection; determine, by means of said processor, for each link, whether the blocking probability corresponding to each of the connections using said link is less than the threshold value corresponding to said connection; increase by 1 the number of wavelengths corresponding to a link if there is a connection using that link where the blocking probability is greater than its corresponding threshold value; and

iteratively repeat the previous three steps.

3. The method of claim 2, CHARACTERIZED because it additionally comprises storing the number of wavelengths required for each connection of said set of connections.

4. The method of claim 2, CHARACTERIZED because it additionally comprises storing the number of wavelengths required for each link of said optical network.

5. The method of claim 1, CHARACTERIZED in that said set of routes is determined using the Dijkstra algorithm.

6. The method of claim 1, CHARACTERIZED because, to determine said probability of blocking, said processor executes the steps of:

determining, for each link in said optical network, the connections using said link; define, for each connection using said link, a first numerical value, t _{0N c} , corresponding to the time in which said connection is finds active and a second numerical value, t _{0FF c} , corresponding to the time in which said connection is inactive; define a plurality of BL _{^ values corresponding to the blocking probability of the connection on the link c in layer I w; and - obtaining a value BP _C corresponding to the probability of blocking of connection c in the network from said BL values.

7. The method of claim 6, wherein said steps of defining a plurality of values _{y ^ BÜ obtain a value BP _C are performed iteratively.